External debt and policy controversy in Korea.

Cho, Jae Ho

I. Introduction

Korea has a reputation for being one of the fastest developing countries in the world, experiencing rapid growth since 1963. Real GNP grew at an average annual rate of 9.0 percent between 1963-93. As a result, Korea's status changed from an underdeveloped country in the 1960s to an upper mid-level developing country in the 1990s. It is well known that foreign debt and the government's active economic policy played an important role in her economic growth.

In addition, the country has a dramatic history of external debt problems. In the early 1980s, Korea was one of the largest debtor nations in the world. Today many economists have come to the consensus that Korea's external debt is no longer a major problem because during the late 1980s the country experienced current account surpluses.

In 1987, in the period of current account surplus, the Korean government revised a schedule for repaying existing debt, and started to repay the debt intensively. At the time, there were different opinions about dealing with the existing current account surpluses. Balassa and Williamson [1] argued that the Korean government should implement an adjustment policy by reducing the current account surpluses. On the other hand, Dombusch and Park [4] did not consider an adjustment policy to be necessary.

In the mid 1990s, the favorable external conditions including a strong yen continue, and the Korean government is demanding a set of structural reforms prior to the establishment of the WTO (World Trade Organization). This paper discusses in detail the policy controversy developed in 1987. With the lesson from the controversy we could get a useful policy direction for the Korean economy in the period of the WTO. This paper also provides the policy implication proposed by the Korean government for those who want to understand the Korean economy and for the policy makers who want to influence it.

This paper consists of four sections. Section II reviews the effect of favorable conditions on the Korean economy in the period of 1980s, and introduces the policy controversy among Balassa-Williamson, Dornbusch-Park, and the Korean government in its dealing with the current surpluses. To evaluate the policy controversy, section III establishes the neoclassical growth model. After investigating the nature of the model, a simulation for the Korean economy is conducted. Section IV concludes with a summary of the Balassa-Williamson and the Dornbusch-Park positions.

[TABULAR DATA FOR TABLE I OMITTED]

II. The Impact of the "Three Lows" and Policy Controversy

Korea's successful pursuit of an export-led development strategy has achieved rapid economic growth and industrialization since the early 1960s. Even though Korea's economic performance is striking compared to other developing countries, in the 1970s the economy was coupled with unfavorable factors such as two oil shocks, political uncertainty, massive investment on heavy industries, and recession in the world economy. Theses adverse effects slackened growth and led to increasing foreign borrowing in the early 1980s. At that time, it seemed that the country might have overborrowed by international standards.

Through successful change in policies in the early 1980s, the country restored growth, lowered inflation, and reduced the current account deficit. In addition, outward economic conditions became more favorable in the mid 1980s because of the decline in oil prices, the decline in world interest rate, and the appreciation of the Japanese yen, all of which are often called the "three lows." Table I shows the trends of the "three lows." The individual impact of the "three lows" on the trade performance is the followings.

The Appreciation of the Japanese Yen

The strong yen had mixed effects on Korea's current account. A positive effect was that the sharp appreciation of the yen against the U.S. dollar increased the price competitiveness of major Korean exports such as cars and electronics items in the U.S. market. On the other hand, the strong yen increased the import price of capital goods from Japan and in turn had adverse impacts on the current account. Such mixed effects were characteristic of the appreciation of the Japanese yen. It is clear that the effect on balance of payments in the mid 1980s was positive.

The Fall in the Oil Price

Korea relies entirely on the import of crude oil. Payment for imported crude oil has been a great burden on the country's development. In fact, crude oil was one of the leading import items during the development period. In the early 1980s the ratio of payment for oil to total imports averaged about 25 percent. After the oil price dropped from $34.1 per barrel in 1981 to about $13.9 in 1988, its ratio was reduced to less than 10 percent even though oil consumption rose. As a result, the 15 percent residual resulted in reducing the economic cost, leading to reduced price levels. The falling oil price level also improved price competitiveness of export products, generating an increase in the real GNP. The positive effects of the fall in oil price were partially offset by import increases occasioned by the rising real GNP. The mixed effects can be examined by evaluating the decomposition of the current account. By applying this method, the author found that falling oil prices had the strongest positive impact on the current account surplus among the "three lows."

The Decline in the World Interest Rate

The world interest rate fell from 16.5 percent in 1981 to 6.8 percent in 1986, which was good news for a heavily indebted country like Korea. The decline in the interest rates reduced the debt burden of the Korean government and of Korean banks. The lower burden of public debt improved the budget balance, permitting debt repayment; it may also have led to increased public consumption and investment expenditures. The lower burden of private debt also increased the availability of funds for investment purposes.

With the help of the "three lows," Korea posted its first current account surplus in 1986, which amounted to some 4.9 percent of GNP. The economic growth rate was recorded at 11.9%, 12.3% and 12.0% in 1986, 1987, and 1988 respectively. The consumer price index almost stabilized at the 2-3 percent inflation in that period. Because of its high economic performance, as reported in Table I, Korea's net foreign debt declined in absolute terms.

In the late 1980s, given the large current account surplus, government policy regarding management of external debt concentrated on repayment of external debt. A number of proposed loans were rejected by the government. Furthermore, the schedule for repaying existing debt was revised. The original target for paying off outstanding debt in the Sixth Five Year Plan (1987-91), established in 1986, was to owe $13.5 billion to foreign investors in 1991. In 1987, during the period of current account surplus, this target was revised to result in a net credit position by the end of 1991. This policy would be natural for a heavily indebted country like Korea. But there was some disagreement about such a presumably impatient plan. Balassa and Williamson [1, 45] argued that such a rapid rundown in debt might incur a misallocation of resources, and suggested that the country should increase investment and consumption spending while keeping a modest current account deficit of 1 or 2 billion dollars because the marginal product of capital far exceeded the world interest rate. They also suggested that the country should take advantage of currently favorable external conditions to start the adjustment process; resulting in much structural strength for the Korean economy without going into recession. In addition, they believed that Korea was overshooting the target, and that it should continue to pursue the sustained growth policy for the long run while keeping a modest current account deficit.

On the other hand, Dornbusch and Park [4] agreed that considering the prevailing uncertainty in the word and domestic labor market development, the large Korean current account surplus might not last long, and that the "three lows" would disappear soon. Therefore, it seemed that the dramatic government action to eliminate the surplus would be premature. They thought that such a large current account surplus was a temporary phenomenon and it might incur trade conflict with developed countries. Accordingly, they suggested that the government trim the current account surplus to the point of possibly evading U.S. trade friction. However, they stressed much more emphasis on a bilateral adjustment through a rapid increase in imports rather than on a broad-based policy as Balassa and Williamson suggested.

According to the policy argument, the critical question was if the government plan for being in a creditor position by the end of 1991 was an optimal strategy for the economy. It is clear that marginal productivity exceeded the world interest rate in Korea in the 1980s.(1) But this was not sufficient to solve the controversy. The answer should be obtained formally. In the following section, a general dynamic equilibrium approach will be presented. The dynamic model is used to describe and simulate the current situation in the Korean economy which will provide a clearer interpretation of the policy controversy.

III. A Dynamic Model and Simulation Results

The concept of current account reflects an intertemporal rather than a static decision on levels of savings and investment. Therefore a model for the policy argument should be dynamic and relevant to the Korean Economy. A dynamic model based upon Blanchard and Fischer [2] will be developed in this section.

The Dynamic Model

The model for a small economy consists of dynamic budget constraints and a lifetime utility function. The economy is composed of a single, infinitely-lived representative economic agent. The economic agent maximizes the utility function under the budget constraint and produces a single good by means of a constant returns technology. The production of a good is divided into consumption, investment spending, net export, and net interest payments on existing debt. The current account deficit denoted by the change in debt, [Mathematical Expression Omitted], is equal to net interest payment minus net exports of goods. Investment spending is the sum of the investment itself, i(t), plus the adjustment cost, [Mathematical Expression Omitted]. The adjustment cost function z(,) is linearly homogeneous in investment goods and capital stock. The rate of capital depreciation ([Delta]) is positive. The world interest rate is assumed to be constant. Repudiation risk, potential liquidity constraints, and insolvency risk are assumed away. Based upon these conditions, the agent determines optimal consumption and investment subjected to dynamic budget constraints. The agent's optimization problem is given by,

[Mathematical Expression Omitted]

where [Rho] is the agent's rate of time preference, f([k.sub.t]) is a neo-classical production function, and [Tau] is a shift parameter.

The first order conditions can be easily derived.(2) With slight modifications of the first order conditions and budget constraints, a dynamic system can be reduced into four differential equations.

[Mathematical Expression Omitted]

[Mathematical Expression Omitted]

[Mathematical Expression Omitted]

[Mathematical Expression Omitted]

The existence and uniqueness of the steady state are ensured by the assumptions already made regarding the utility (concave) and budget constant (convex). An imposer feature of this system is that it is "block recursive," implying that the dynamics of (q, k) are independent of shocks originating from consumption - current account dynamics. At a neighborhood around the equilibrium, the dynamics of (q, k) can be rewritten as,

[Mathematical Expression Omitted]

Since the determinant of the sub matrix for [Mathematical Expression Omitted], is negative, one negative eigenvalue, denoted by [Theta], is such that

[Mathematical Expression Omitted]

where [Delta], r, [Tau], [Psi][prime](0), [Mathematical Expression Omitted], [Mathematical Expression Omitted], and [Mathematical Expression Omitted] represents the steady stare value of capital. Thus, the long run equilibrium is a saddle point.

Assuming that the economy starts out with initial stocks ([k.sub.0], [b.sub.0]), the stable dynamic paths followed by k and q are given by(3)

[Mathematical Expression Omitted]

[Mathematical Expression Omitted]

Finely, the steady state values me given by(4)

[Mathematical Expression Omitted]

Dynamic Effects of Change in Parameters

One can analyze the dynamic effects of changes in the parameters in the model. This section places emphasis on the change in two key parameters: a shift parameter of productivity level, [Tau], and the world interest rate, r. As discussed later, the effects of these two parameters are closely related to that of the "three lows" on the Korean economy in the mid 1980s. These effects should be also distinguished between a temporary and a permanent one because Dornbusch and Park considered the "three lows" as a temporary phenomenon while Balassa and Williamson as a permanent phenomenon. Depending on the duration of the impacts, the response of the economy is different.

Response to Permanent Shocks. A permanent increase in productivity level ([Tau]) and a drop in interest rates (r) increase the new steady state value of capital and change the value of the characteristic root, which, when the changes occur, lead to the upward jumps in the shadow price of capital (q), investment (i), output (f(k)), and consumption (c) instantaneously (and permanently). For the formal proofs for this, the reader is referred to Cho [3].

Response to Temporary Shocks. Suppose that at time (t = 0), [Tau] increases and r decreases, but both are expected to be restored to their original levels at time (t = T). As soon as the shocks occur, the shadow price (q) and investment (i) will jump instantaneously (and temporarily). As a result of the initial rise in (q) and (i), capital and external debt begin to accumulate analogous to reasons noted in connection with the permanent shocks. At the same time consumption will increase by precisely the same amount as if the shocks were permanent. The accumulation of capital and debt will serve as initial conditions for the dynamic beyond time (t = T) when the shocks revert permanently to their original level. They then will determine the new steady state of equilibrium. With no new information being received at the time (t = T), the control variables will not jump any more.

The "Three Lows," Openness Policy, and Output Function

Given the dynamic system established above, we can return to the policy controversy developed in 1987. As discussed previously, the economy had quickly overcome the debt problem and had recorded current account surpluses in the mid 1980s. Clearly, the recovery of the economy was mainly due to two factors: favorable external conditions such as the "three lows" and the outward oriented policy sustained since the 1960s. The outward oriented policy under the "three lows" had such a great impact on economic expansion that the economy recovered from the ill effects caused by the domestic and the foreign recessions of the 1970s.

The essence of the policy controversy is about whether or not it was optimal for Korea to reduce its foreign debt levels. To answer the question under the established framework, one could introduce the impacts of the "three lows" on the outward oriented economy into the model. Then the remaining questions are why and how the "three lows" and the outward oriented policy contribute to an increase in productivity and economic growth in the period of the "three lows." For these questions, we need to review the recent development theory briefly.

Recently much of the literature has used cross country regression to search for empirical linkages between long run growth rates and a variety of openness variables. For this study, many economists have developed exogenous and endogenous growth models as well, but qualitatively the derived results are likely to be the same. They find a positive and robust relationship between growth and openness variables. Specifically, a study developed by Romer [6] argues that the theoretical ties between growth and trade seem to run through improved productivity. This study implies that openness to trade results not only in raising output but also in improving the marginal product of capital through creation of new investment opportunities. In this sense, openness to trade and new investment spending become complementary factors of production, determining the productivity of the economy. Absence of complementarity would mean a lower rate of growth of productivity or marginal product of capital, hence lower investment incentives with high export opportunities and expansion, a result that is not consistent with existing empirical evidence.

Based upon this theoretical background in which most of the theoretical research has explicitly focused on the Asian miracles including Korea, it is not difficult to infer that the effect of more variation in Korea's trade performance engendered by the "three lows" is thought to be related to the variation of the technological factor.

Returning to the model for the analysis of these effects, the production function in the above model can be modified. Under the assumption that the "three lows" significantly affected Korea's economic growth and productivity through the increase in trade performance, the reduced form can be disentangled by introducing the price (terms of trade) effect on the production function in the model. The world interest rate is already included in the model.

Given a simple model in which the price effect on output is included, the output function in terms of labor in efficiency units, [y.sub.t] [equivalent to] [Y.sub.t]/[L.sub.t] exp([Lambda]t), is derived as

[Mathematical Expression Omitted],

where A is a constant, [P.sub.x] and [P.sub.n] are the price of a tradable good and imported oil in terms of the price of import, and [Alpha], [Beta] are the share of capital and oil import respectively. The derivation of Korea's net output function is in Appendix A.

Concerning the effect of the "three lows" on the economy, it is important to describe the [Tau] function using parameter values. Given terms of trade and other parameters values (see the next section), the trend of the [Tau] function is shown in Figure 1. It has been continuously increasing from 1986. The trend is consistent with the assumption of productivity improvement during "the three lows" period. Since [Alpha] + [Beta] [less than] 1, the net output function has a concave shape consistent with the assumption made in the model in the previous section.

Simulation Results

The model is used to simulate the Korean economy from 1981 to 1992 to evaluate the policy controversy. The historical simulation is carried out respectively for two periods: before "the three lows" period (1981-85) and during "the three lows" period (1986-92). Given actual parameter values in Table II, the simulation for the period before "the three lows" will be calibrated to obtain other appropriate parameter values which cannot be directly obtained from the data. Then using these parameter values as well as the new values of the productivity and the world interest rates after the shocks, the model will simulate the economic performance for the period during the "three lows." The simulation for this period is divided into two cases, i.e., the cases of permanent shock and temporary shock. The duration of the temporary shock is assumed to have lasted for 4 years - from 1986 to 1989. This assumption is based upon the observation that the [Tau] function continuously increased during that period and then slightly declined from 1990. The duration of the permanent shock is assumed to have continued from 1986 for the entire period of the simulation.

For a simulation, the structure of the model and parameter values can be specified. The utility function is assumed to be logarithmic and u([C.sub.t], [L.sub.t]) = [L.sub.t] ln([C.sub.t], [L.sub.t]). The investment function [TABULAR DATA FOR TABLE II OMITTED] is of a familiar form [I.sub.t][1 + [Phi]([I.sub.t]/[K.sub.t])], where [Phi] is a coefficient of adjustment cost and is linear. Also, all variables are divided by labor in efficiency units ([y.sub.t] [equivalent to] [Y.sub.t]/[L.sub.t] exp([Lambda]t)) rather than labor units, where [Lambda] is the Harrod neutral technological progress. It can also easily be shown that the modified model has the same characteristic of dynamics as the previous one in section III. The only differences are in the form of expression. The equations for simulations and methods are described in Appendix B.

There are several parameters in the model. The first, the world interest rates (r) in the model is defined as the interest rate paid by Korea. The average of the actual interest rate paid by Korea during 1981-85 is 9.0%. In the case of the permanent shock since 1986, 7% is assumed to hold for the entire period of simulation. In the temporary case, it is assumed that an interest rate of 7% continues during 1986-89 and then increases to 8.0% in 1990. The value of [Tau] in equation (6), as shown in Figure 1, is calculated by 0.58 in 1985 which increases to 0.70 in 1986 because of the "three lows." In the permanent case, this value is assumed to maintain its level for the entire period of simulation. In the temporary case, it is assumed to be 0.70 during 1986-89 and then decreases to 0.64.

The coefficient of the adjustment cost function ([Phi]) and the rate of technological progress ([Lambda]) are chosen to be 1.5 and 4.5% respectively based upon the results of the calibration of actual and optimal investments and consumption. These values are assumed to be the same for the entire period of the simulation. The growth rate of employee population (n) and the rate of depreciation are given to be 2.0% and 5%. Capital share ([Pi] = [Alpha]/(1 - [Beta])) and constant term (A) in equation (6) are estimated to be 0.31 and 2.0 respectively.(5) These values are obtained through the estimation of the production function. Finally, in equation (2) for the characteristic root ([Theta]), [Psi][prime](0) is chosen to be 0.06 in consideration of optimal investment and actual investment.

The values of parameter r and [Tau] are summarized in Table II. In another empirical study of the Korean economy, Park and An [5] assumed the values of capital share, adjustment cost coefficient, the rate of technological progress, and the rate of depreciation are assumed to be 0.35, 1.5, 4%, and 5% respectively. These values are similar to those in this paper. Given the structure of the dynamic model and the parameter values, the simulation for the Korean economy before the "three lows" (1981-85) can be carried out as illustrated in Figure 2 and Figure 3(a). Before the "three lows," the simulation results in the root mean square percent errors of investment, capital stock, GDP, consumption, and debt to be 3.5%, 2.2%, 3.8%, 3.5% and 4.3% respectively. Therefore, assuming the above parameters, the simulated data before the "three lows" trace the actual data quite accurately.(6)

Response to Permanent Shocks. During the "three lows," the simulation is conducted in both the temporary and the permanent cases. Given the parameter values in Table II, the initial levels of ([Mathematical Expression Omitted], [Theta]) are calculated using the steady state condition and equation (2) (For details, see Appendix B.) From the calculation, the steady state level of capital based upon the 1985 parameter's values is 10.42 in efficiency labor units and it increases to 17.13. The characteristic root decreases from -0.064 to -0.071. The new steady state levels of ([Mathematical Expression Omitted], [Theta]) in turn determine the new optimal paths of control variables. Given the new values of steady state capital stock and characteristic root after the positive permanent shocks, the sizes of the jump of control variables are calculated as follows. The initial levels of the variables are derived in Appendix B. Investment spending calculated by labor in efficiency units in 1985 is 1.14; it then rises to 1.63 in 1986. By the same token, [q.sub.0] is calculated as 1.18 in 1985 and it then jumps to 1.48 in 1986. Also the shift in consumption is calculated from 2.90 in 1985 to 3.30 in 1986. Finally, the simulated level of GDP due to the permanent shocks jumps from 4.39 in 1985 to 5.34 in 1986. These results are summarized in Table III.

[TABULAR DATA FOR TABLE III OMITTED]

Response to Temporary Shocks. Using the same procedure as the permanent case, the new steady state level of capital and the characteristic root are calculated as 12.74 in efficiency labor unit and -0.066 respectively. The size of the jump of the control variables are calculated in Table III.

Switching the unit from an efficiency unit to an actual value, the shift in optimal investment spending, capital stock, and GDP are depicted in Figure 2 (The units in x-axis and y-axis are year and trillion won respectively). In the permanent case, the transitional dynamics in Figure 2 show higher levels of simulated data, denoted by a line, for GDP, investment, and capital stock than do the actual data, denoted by a dotted line, in the late 1980s.

The trends of dynamic paths of the economic variables are consistent with the results derived in section III. Specifically, for the period 1986-89, the simulated path of investment exceeds the actual path of investment, and the trend of investment has reversed for 1990-92. During the period of 1986-89, the path showed that investment spending was not sufficiently raised in the period of the high level of productivity. Instead the government concentrated on the repayment of external debts. So this policy resulted in a reduction of the private firm's motivation for new investment. The growth rates of equipment investment (the machinery purchases among equipment investment) in 1986, 1987, and 1989 were 19.0% (5.6 trillion won), 13.0% (6.0 trillion won), and 15.2% (8.9 trillion won). The levels of the machinery purchase contrasted with the 18.3 trillion won in 1990, and 20.2 trillion won in 1991 when the government started adjustment policy. In comparison, in the period of 1976-78 when there was a world recession, the average growth rate of the equipment investment was 40%. These facts thus imply that investment spending including manufacturing did not increase as much as the domestic productivity improvement as a result of the "three lows."

During the period of 1990-92, the trend has reversed i.e., the actual investment exceeds the simulated. This was mainly caused by the early 1990s overinvestment in construction due to real estate speculation. According to the data, the growth rate of construction investment in 1986 was 3.1 percent, which rose to 29.1 percent in 1990, and in a year the investment of housing grew 61.5% in 1989. As a result, construction investment in the early 1990s far exceeded equipment investment. This overinvestment in construction was not caused by the improvement in productivity but was caused by the skyrocketing increase of capital gain in the real estate market. This phenomenon originated in the misallocation of resources, pointed out by Balassa and Williamson [1]. The Korean government's mismanagement of its current account surplus resulted in the construction boom in the early 1990s.

In the temporary case, the transitional dynamics in Figure 2 show lower levels of simulated data, denoted by a long dotted line, for consumption, investment, GDP, and capital stock than do the actual data in the late 1980s. These results imply that the temporary case does not fit well with the actual data.

Finally, as depicted in Figure 3(a) the simulated level of net foreign debt, deflated by the GDP deflator in terms of the won currency, increased for the first part of the boom period and then declined slightly. The large gap between the optimal and actual debt was caused by the government's repayment of the external debt during that period. Furthermore, in the early 1990s the simulated data for debt, in the temporary case, tends to rise as the "three lows" are assumed to have disappeared at the end of 1989.

One can also examine the trend of the ratio of net debt to GDP. The upper line in Figure 3(b) which declines, represents the trend of the ratio of debt to GDP when the actual level of consumption is used in the simulation. It implies that if the Korean economy had followed an optimal path of investment in 1986-89 given the actual level of consumption, the ratio of net debt to GDP eventually could have decreased without active government intervention against the external debt problem. This means that the economy could have an increase in structural strengthening if the economy had higher levels of investment spending by taking advantages of the improvement in productivity during the "three lows" period. This view confirms the Balassa and Williamson position. On the other hand, Dornbusch and Park [4, 433] argued that "Korea's investment rate is more than 30 percent of GNP. There is little to suggest that capital imports are necessary because capital is in short supply." This view is neither supported by the result of the simulation nor by the recent study concerning the correlation amongst the openness policy, output growth, and productivity. Even during 1977-79, when Korea suffered from unfavorable external conditions, the average investment rate was recorded at 32.3%, which exceeded the investment rate of 31.3% during 1987-89, when favorable conditions were apparent. Furthermore, the Korean government and Dornbusch-Park's positions were inconsistent with the export-led economic policy which has been pursued since the 1960s.

In the early 1990s, the government had realized this and addressed this with the adjustment policy. But the economic circumstance became worse because of the increasing pressure of trade openness and appreciation of won from the developed countries. Furthermore, the effect of the "three lows" was reduced as time passed. This ill-timed policy and the adverse effect of the debt repayment policy suppressed economic growth, and current account recorded in deficits in the early 1990s. In 1992, the current account deficit was 4.6 billion dollars, and the GNP growth rate was 5.0 percent, the lowest (except 1980) over the developing period.

IV. Conclusion

This paper has analyzed the performance of the Korean economy focusing on external debt, economic growth, and the pertinence of the Balassa-Williamson position. In conclusion, according to the simulation results and other evidences, a part of the improvement in productivity during the boom period was likely to be temporary. There was nevertheless a sufficiently permanent component to justify adjustment of the surplus developed by 1987. Thus, if the Korean government had accepted Balassa and Williamson's suggestion during the boom period, the Korean economy would have maintained a higher level of economic growth with an optimal level of external debt. In this context, the revised Sixth Five Year Plan's proposal that Korea be in a creditor position in 1991 is not a well established plan.

Through the policy controversy, we can review the characteristics of the Korean Economy and then suggest a policy direction based upon the lesson from it. Korean economic growth is mainly led by export expansion. Since export expansion depends on the world's economic situation, the Korean economy has fluctuated with the word economy. Rapid but sharply fluctuating growth has been a prominent characteristic of Korean development. Therefore, one can call the Korean economy a "bicycle economy." To run a "bicycle economy" properly, it is necessary to make rules for economic policies.

First, to maintain its speed, the country should continue its export promotion policy through globalization of the economy. This strategy is consistent with the policy pursued since the 1960s. To increase the volume of export, Korean firms and the government should promote the upgrading of Korea's industries toward continued product specialization in favor of higher valued-added goods. These efforts are only possible by raising the level of technology. Second, to maintain its balance, the country needs to implement an adjustment policy concentrated on increasing domestic demand. A policy for increasing domestic spending should not only cause the stability of the economy but also provide a base for domestic industries, raising their competitive capability in the world market. The larger the domestic demand, the more the reduction in the fluctuation of the Korean economy due to external shocks. Also, the industries that survive in the domestic market can compete well in the hostile world economic environment.

The necessity of the openness and the globalization of the Korean economy are consistent with the ideas of the WTO (World Trade Organization) which prepares to start in 1995. On the departure of the new world trade system, Korea is demanding a set of adjustment policies to adjust to the hostile external environment. The Korean economy should consider the new trade system as another opportunity, which is a different economic environment, to implement an adjustment policy to give structural strength for the Korean economy. As experienced in the late 1980s, if the government successfully adjusts to the new trade system in an active manner rather than a passive one, it will improve its productivity and economic growth. By doing so, the Korean economy can put one more step towards the level of a developed economy.

Appendix A

The purpose of the following simple model is to disentangle the parameter, [Tau], of Korea's output function in order to include the effect of price changes on the output function. Assume the following circumstances: (1) the economy produces an export good (X) which is exchanged in the world market for a consumption good (M). (2) The price of a tradable good and imported oil in terms of the price of M, [P.sub.x] and [P.sub.n], respectively, are beyond the control of the economy. (3) Three inputs, capital (K), labor (L), and imported oil (N), are used in the production of (X): X = G(K, N, Lexp([Lambda]t)). (4) Net output (output minus the cost of imported oil) should be maximized.

Given these assumptions, a simple model is:

Max [p.sub.x]G(K, N, L exp([Lambda]t)) - [p.sub.n]N.

To maximize the net output with respect to N, the first order condition should be satisfied

[Delta]G/[Delta]N = [p.sub.n]/[p.sub.x].

Net output function is defined by

[Y.Sub.T] [equivalent to] [P.sub.x][G(K, N, L exp([Lambda]t)) - N[Delta]G/[Delta]N]. (1)

The first order condition becomes [Delta]K/[Delta]N - [Beta]A[K.sup.[Alpha]][N.sup.[Beta]-1] [(L exp([Lambda]t)).sup.[Gamma]] = [p.sub.n]/[p.sub.x], and the optimal choice of N is derived by

[N.sup.[Beta]] = [([Beta][p.sub.x]/[p.sub.n]).sup.[Beta]/(1-[Beta])][(L exp([Lambda]t)).sup.[Beta][Gamma]/(1-[Beta])][K.sup.[Alpha][Beta]/(1-[Beta ])]. (2)

Substituting (2) into (1), then

[Y.sub.t] = (1 -[Beta])[p.sub.x]([Beta][p.sub.x][([Beta][p.sub.x]/[p.sub.n]).sup.[Beta]/( 1-[Beta])] [A.sup.1/(1-[Beta])][(L exp([Gamma]t)).sup.[Gamma]/(1-[Beta])][K.sup.[Alpha][Beta]/(1-[Beta])]

[equivalent to] [Tau]([Beta],[p.sub.x],[p.sub.n])[A.sup.1/(1-[Beta])][(L exp([Lambda]t)).sup.[Gamma]/(1-[Beta])][K.sup.[Beta]/(1-[Beta])].

From the equation, [Tau] is defined as (1 - [Beta])[p.sub.x][([Beta][p.sub.x]/[p.sub.n]).sup.[Beta]/(1-[Beta])]. Then given the condition of [Alpha] + [Beta] + [Gamma] = 1, the output function labor in efficiency units, [y.sub.t] [equivalent to] [Y.sub.t]/[L.sub.t] exp([Lambda]t), is represented by [y.sub.t] = [Tau]([p.sub.x], [p.sub.n], [Beta])[A.sup.1/(1-[Beta])] [Mathematical Expression Omitted].

Appendix B

The following equations are used for simulations.

Debt: [Mathematical Expression Omitted]

Capital accumulation function: [Mathematical Expression Omitted]

Investment function: [Mathematical Expression Omitted]

Consumption function: [Mathematical Expression Omitted]

Output function: [Mathematical Expression Omitted].

The investment function is derived by combining the marginal cost of investment and the present value of the marginal product (see footnote 2). If r = p + [Lambda], consumption in efficiency units is constant so that, along the optimal path, consumption per capita grows at the rate of technological progress ([Lambda]). This assumption is used here for the simulation of consumption. Finally the initial levels of the control variables in the dynamic model are given by:

consumption: [Mathematical Expression Omitted]

the shadow price of capital: [Mathematical Expression Omitted]

investment: [Mathematical Expression Omitted].

(The initial level of investment is derived based upon the following equations:

[Mathematical Expression Omitted]

The steady state level of capital [Mathematical Expression Omitted] is obtained by the steady state condition in equation (5): [Mathematical Expression Omitted], so it holds [Mathematical Expression Omitted], where [pi] = [Alpha]/(1 - [Beta]), [A.sup.*] = [A.sup.1/(1-[Beta])]. Finally the path of the marginal product of capital is determined by [Mathematical Expression Omitted].

Earlier versions of this paper were presented at the celebration meeting for professor Duk Choong Kim's birth held in Seoul in 1994, and at the Southern Economic Associations's annual conference in Orlando, Florida in November of 1994. The author wishes to thank professor Bela Balassa, Carl Christ, Enrica Detragiache, and an anonymous referee for their helpful comments. The financial support from the chief director of the University of Ulsan is greatly acknowledged.

1. According to the author's calculation using the values of the parameters in section III, the marginal product of capital in 1985 was 16.4 percent, while the interest rate paid by the Korean economy was 9.5 percent.

2. The first conditions for investment and consumption are derived by [Mathematical Expression Omitted], [Mathematical Expression Omitted], [Mathematical Expression Omitted] where [Epsilon] = -u[prime]/u[double prime][c.sub.t] [center dot] [q.sub.t] is the shadow price of capital. The dynamic of capital stock can be expressed in terms of [Mathematical Expression Omitted] where [q.sub.t] [greater than or equal to] 1, it holds it [Mathematical Expression Omitted]. The solvency condition is given as [Mathematical Expression Omitted].

3. Equation (4) is obtained as follows. [Mathematical Expression Omitted].

4. Under the assumption of ([Rho] = r), four differrential equations reduce to three differential equations with four unknown variables: ([Mathematical Expression Omitted], [Mathematical Expression Omitted], [Mathematical Expression Omitted], [Mathematical Expression Omitted]). If [Mathematical Expression Omitted] is given as a parameter such as the initial consumption level ([c.sub.0]), the steady state of ([Mathematical Expression Omitted], [Mathematical Expression Omitted], [Mathematical Expression Omitted]) are then determined by the given ([c.sub.0]) and ([Mathematical Expression Omitted], [Mathematical Expression Omitted], [Mathematical Expression Omitted]). (If [Rho] [not equal to] r, consumption should be zero in the steady state because the equation for consumption becomes [Mathematical Expression Omitted]. This is not a reasonable result.)

5. The production function between 1974-92 where (N) is oil input is estimated.

[Mathematical Expression Omitted]

([R.sup.2] = 0.68, number of observations are 17; the coefficient standard errors are in parentheses)

6. The investigation of robustness of economic variables during the "three lows" period of 1987-92 is not necessary. The reason is that the simulated results over that period are derived on the assumption that the Korean government implemented the adjustment policy optimally.

References

1. Balassa, Bela and John Williamson. Adjusting to Success, Balance of Payment Policy in the East Asian NICs. Washington D.C.: Institute for International Economics, 1987.

2. Blanchard, Olivier J. and Stanley Fischer. Lectures on Macroeconomics. Cambridge, Mass.: The MIT press, 1989.

3. Cho, Jae Ho. "External Debt and Openness Policy in Korea," Ph.D. dissertation, The Johns Hopkins University, 1992.

4. Dornbusch, R. and Y. C. Park, "Korean Growth Policy." Brookings Papers on Economic Activity, 2, 1987.

5. Park, W. A. and J. H. An. "Savings, Investment and the External Debt in Korea." Korea Development Institute Working Paper, no. 8812, 1988.

6. Romer, Paul M. "Cross Country Determinations of Growth and Technological Change." The World Bank Working Paper, 1989.